OpenAI delays the release of its open model, again

OpenAI, a frontrunner in artificial intelligence development, has once again pushed back the release of its highly anticipated open model. OpenAI CEO Sam Altman announced on Friday that the model, which had already seen a month-long delay earlier this summer, is now indefinitely postponed for further safety testing. This news comes as a significant update for the AI community and enthusiasts, including visitors to AI tool listing websites like Proaitools.

Altman conveyed the company’s position in a post on X (formerly Twitter), stating, “We need time to run additional safety tests and review high-risk areas. we are not yet sure how long it will take us.” He emphasized the irreversible nature of releasing weights into the public domain: “While we trust the community will build great things with this model, once weights are out, they can’t be pulled back. This is new for us and we want to get it right.” This cautious approach highlights OpenAI’s commitment to responsible AI deployment.

The release of OpenAI’s open model stands as one of the most keenly awaited AI developments of the summer, alongside the expected launch of GPT-5. Unlike GPT-5, the open model is designed to be freely downloadable and runnable locally by developers, promising a new era of innovation. Through these strategic releases, OpenAI aims to reaffirm its position as a leading AI research lab in Silicon Valley, navigating an increasingly competitive landscape with major players like xAI, Google DeepMind, and Anthropic investing heavily in their own AI endeavors.

For developers keen to experiment with OpenAI’s first open model in years, the delay means an extended wait. Reports from TechCrunch indicate that this open model is expected to offer reasoning capabilities comparable to OpenAI’s ‘o-series’ models, with the ambition of setting a new ‘best-in-class’ standard among other open-source alternatives.

The open AI model ecosystem has recently seen heightened competition. Just last Friday, Chinese AI startup Moonshot AI unveiled Kimi K2, a formidable one-trillion-parameter open AI model that has already demonstrated superior performance over OpenAI’s GPT-4.1 AI model in several agentic-coding benchmarks, underscoring the rapid advancements across the global AI landscape.

Reflecting on an earlier delay in June, Altman had alluded to an “unexpected and quite amazing” achievement without divulging specifics. Aidan Clark, OpenAI’s VP of research and leader of the open model team, reiterated the company’s sentiment on X: “Capability wise, we think the model is phenomenal — but our bar for an open source model is high and we think we need some more time to make sure we’re releasing a model we’re proud of along every axis.”

While previous reports by TechCrunch suggested that OpenAI was exploring features for its open AI model to connect with its cloud-hosted AI models for more complex queries, it remains uncertain whether these advanced capabilities will be integrated into the final release.